化工过程中虚假宽松约束的修正预测控制

过程控制中广泛存在着对变量的各种约束条件,因此需要设计能够满足约束条件的控制策略。约束预测控制能够处理各种输入和输出约束,是一种有效的控制方法。但在实际过程控制中,由于预测控制不检测下层回路的实际执行信息,当控制作用超出系统实际最大能力而不能完全下载时,标称约束成为虚假宽松约束,导致系统响应偏离预期,约束条件无法满足。该文基于线性系统离散状态空间模型,讨论有输入和输出区间约束条件的基于区间控制思想的预测控制。为保证虚假宽松约束发生时预测控制仍能够正确有效实施,提出及时进行一系列预测模型数据和约束条件的修正算法,并使约束条件得到满足。利用绝热连续搅拌釜式反应器(CSTR)模型的控制仿真实验证明了该算法的有效性。

Predictive control of chemical processes with false and loose constraints

Process controllers need to fulfil all kinds of variable constraints that commonly exist in process control.Model predictive control(MPC) can efficiently deal with constraints on input and output variables.Model predictive control does not normally measure information in local control loops.If the control action is beyond the normal system capability and cannot be controlled,the nominal constraints become false and loose constraints,the practical system response deviates from the expected output,and the constraints cannot be satisfied.This paper describes an MPC based on zone control,a discrete-time state-space model of a linear system,with input and output zone constraints.The MPC constraints are met by an update algorithm using a predictive model with constraints.Simulations of the control of a continuously stirred-tank reactor(CSTR) show the effectiveness of this algorithm.